Understanding Operating Systems

Chapter 5
Process Management

Learning Objectives
Several causes of system deadlock
The difference between preventing and avoiding
deadlocks
How to detect and recover from deadlocks
The concept of process starvation and how to detect
and recover from it
The concept of a race and how to prevent it
The difference between deadlock, starvation, and
race
2

Deadlock
Resource sharing
Memory management and processor sharing

Many programs competing for limited resources

Lack of process synchronization consequences
Deadlock: deadly embrace

Two or more jobs placed in HOLD state

Jobs waiting for unavailable vital resource
System comes to standstill
Resolved via external intervention

Starvation
Infinite postponement of job
3

Deadlock (continued)
More serious than starvation
Affects entire system
Affects more than one job
Not just a few programs

All system resources become unavailable

Example: traffic jam (Figure 5.1)

More prevalent in interactive systems
Real-time systems
Deadlocks quickly become critical situations

No simple and immediate solution

4

Deadlock (continued)

Seven Cases of Deadlock

Nonsharable/nonpreemptable resources
Allocated to jobs requiring same type of resources

Resource types locked by competing jobs

File requests
Databases
Dedicated device allocation
Multiple device allocation
Spooling
Disk sharing
Network
6

Case 1: Deadlocks on File Requests

Jobs request and hold files for execution duration
Example (Figure 5.2)
Two programs (P1, P2) and two files (F1, F2)
Deadlock sequence
P1 has access to F1 and also requires F2
P2 has access to F2 and also requires F1

Deadlock remains
Until one program withdrawn or
Until one program forcibly removed and file released

Other programs requiring F1 or F2

Put on hold for duration of situation
7

Case 1: Deadlocks on File Requests

(continued)

Case 2: Deadlocks in Databases

Two processes access and lock database records
Locking
Technique
One user locks out all other users
Users working with database

Three locking levels

Entire database for duration of request
Subsection of database
Individual record until request completed

Case 2: Deadlocks in Databases

(continued)
Example: two processes (P1 and P2)
Each needs to update two records (R1 and R2)
Deadlock sequence

P1 accesses R1 and locks it

P2 accesses R2 and locks it
P1 requests R2 but locked by P2
P2 requests R1 but locked by P1

Race between processes

Results when locking not used
Causes incorrect final version of data
Depends on process execution order
10

Case 2: Deadlocks in Databases

(continued)

11

Case 3: Deadlocks in Dedicated

Device Allocation
Limited number of dedicated devices
Example
Two programs (P1, P2)
Need two tape drives each
Only two tape drives in system

Deadlock sequence

P1 requests tape drive 1 and gets it

P2 requests tape drive 2 and gets it
P1 requests tape drive 2 but blocked
P2 requests tape drive 1 but blocked
12

Case 4: Deadlocks in Multiple

Device Allocation
Several processes request and hold dedicated
devices
Example (Figure 5.4)
Three programs (P1, P2, P3)
Three dedicated devices (tape drive, printer, plotter)
Deadlock sequence

P1 requests and gets tape drive

P2 requests and gets printer
P3 requests and gets the plotter
P1 requests printer but blocked
P2 requests plotter but blocked
P3 requests tape drive but blocked
13

Case 4: Deadlocks in Multiple

Device Allocation (continued)

14

Case 5: Deadlocks in Spooling

Virtual device
Dedicated device made sharable
Example
Printer: high-speed disk device between printer and
CPU

Spooling
Process
Disk accepts output from several users
Acts as temporary storage for output
Output resides in disk area until printer accepts job data
15

Case 5: Deadlocks in Spooling

(continued)
Deadlock sequence
Printer needs all job output before printing begins

Spooling system fills disk space area

No one job has entire print output in spool area
Results in partially completed output for all jobs
Results in deadlock

16

Case 6: Deadlocks in a Network

No network protocols controlling network message
flow
Example (Figure 5.5)
Seven computers on network
Each on different nodes

Direction of arrows
Indicates message flow

Deadlock sequence
All available buffer space fills

17

Case 6: Deadlocks in a Network

(continued)

18

Case 7: Deadlocks in Disk Sharing

Competing processes send conflicting commands
Scenario: disk access

Example (Figure 5.6)

Two processes
Each process waiting for I/O request
One at cylinder 20 and one at cylinder 310

Deadlock sequence
Neither I/O request satisfied
Device puts request on hold while attempting to fulfill
other request for each request

Livelock results
19

Case 7: Deadlocks in Disk Sharing

(continued)

20

Conditions for Deadlock

Four conditions simultaneously occurring prior to
deadlock or livelock

Mutual exclusion
Resource holding
No preemption
Circular wait

All needed by operating system

Must recognize simultaneous occurrence of four
conditions

Resolving deadlock
Removal of one condition
21

Conditions for Deadlock (continued)

Mutual exclusion
Allowing only one process access to dedicated
resource

Resource holding
Holding resource and not releasing it
Waiting for other job to retreat

No preemption
Lack of temporary reallocation of resources

22

Conditions for Deadlock (continued)

Circular wait
Each process involved in impasse
Waiting voluntarily resource release by another so at
least one can continue

All four required for deadlock occurrence

Deadlock remains until one condition removed

23

Modeling Deadlocks
Directed graphs
Circles represent processes
Squares represent resources
Solid arrow from resource to process
Process holding resource

Solid arrow from a process to resource

Process waiting for resource

Arrow direction indicates flow

Cycle in graph
Deadlock involving processes and resources
24

Modeling Deadlocks (continued)

25

Modeling Deadlocks (continued)

Three graph scenarios to help detect deadlocks
System has three processes (P1, P2, P3)
System has three resources (R1, R2, R3)

Scenario one: no deadlock

Resources released before next process request

Scenario two: deadlock

Processes waiting for resource held by another

Scenario three: no deadlock

Resources released before deadlock

26

Modeling Deadlocks (continued)

No deadlock
Resources released before next process request

27

Modeling Deadlocks (continued)

28

Modeling Deadlocks (continued)

Deadlock
Processes waiting for resource held by another

29

Modeling Deadlocks (continued)

30

Modeling Deadlocks (continued)

No deadlock
Resources released before deadlock

31

Modeling Deadlocks (continued)

32

Modeling Deadlocks (continued)

Another example
Resources of same type
Allocated individually or grouped in same process
Graph clusters devices into one entity

Allocated individually or grouped in different process

Graph clusters devices into one entity

33

Modeling Deadlocks (continued)

34

Strategies for Handling Deadlocks

Prevention
Prevent occurrence of one condition
Mutual exclusion, resource holding, no preemption,
circular wait

Avoidance
Avoid deadlock if it becomes probable

Detection
Detect deadlock when it occurs
Recover gracefully

Recovery
Resume system normalcy quickly and gracefully
35

Strategies for Handling Deadlocks

(continued)
Prevention eliminates one of four conditions
Complication: every resource cannot be eliminated
from every condition
Mutual exclusion
Some resources must allocate exclusively
Bypassed if I/O device uses spooling

Resource holding
Bypassed if jobs request every necessary resource at
creation time
Multiprogramming degree significantly decreased
Idle peripheral devices
36

Strategies for Handling Deadlocks

(continued)
Prevention (continued)
No preemption
Bypassed if operating system allowed to deallocate
resources from jobs
Okay if job state easily saved and restored
Not accepted to preempt dedicated I/O device or files
during modification

Circular wait

Bypassed if operating system prevents circle formation

Use hierarchical ordering scheme
Requires jobs to anticipate resource request order
Difficult to satisfy all users
37

Strategies for Handling Deadlocks

(continued)
Avoidance: use if condition cannot be removed
System knows ahead of time
Sequence of requests associated with each active
process

Dijkstras Bankers Algorithm

Regulates resources allocation to avoid deadlock
No customer granted loan exceeding banks total
capital
All customers given maximum credit limit
No customer allowed to borrow over limit
Sum of all loans will not exceed banks total capital
38

Strategies for Handling Deadlocks

(continued)

39

Strategies for Handling Deadlocks

(continued)

40

Strategies for Handling Deadlocks

(continued)

41

Strategies for Handling Deadlocks

(continued)

42

Strategies for Handling Deadlocks

(continued)
Operating systems deadlock avoidance assurances
Never satisfy request if job state moves from safe to
unsafe
Identify job with smallest number of remaining
resources
Number of available resources => number needed for
selected job to complete
Block request jeopardizing safe state

43

Strategies for Handling Deadlocks

(continued)
Problems with the Bankers Algorithm
Jobs must state maximum number needed resources
Requires constant number of total resources for each
class
Number of jobs must remain fixed
Possible high overhead cost incurred
Resources not well utilized
Algorithm assumes worst case

Scheduling suffers
Result of poor utilization
Jobs kept waiting for resource allocation
44

Strategies for Handling Deadlocks

(continued)
Detection: build directed resource graphs
Look for cycles

Algorithm detecting circularity

Executed whenever appropriate

Detection algorithm
Remove process using current resource and not
waiting for one
Remove process waiting for one resource class
Not fully allocated

Go back to step 1
Repeat steps 1 and 2 until all connecting lines removed
45

Strategies for Handling Deadlocks

(continued)

46

Strategies for Handling Deadlocks

(continued)

47

Strategies for Handling Deadlocks

(continued)
Recovery
Deadlock untangled once detected
System returns to normal quickly

All recovery methods have at least one victim

Recovery methods
Terminate every job active in system
Restart jobs from beginning

Terminate only jobs involved in deadlock

Ask users to resubmit jobs

Identify jobs involved in deadlock

Terminate jobs one at a time
48

Strategies for Handling Deadlocks

(continued)
Recovery methods (continued)
Interrupt jobs with record (snapshot) of progress
Select nondeadlocked job
Preempt its resources
Allocate resources to deadlocked process

Stop new jobs from entering system

Allow nondeadlocked jobs to complete
Releases resources when complete
No victim

49

Strategies for Handling Deadlocks

(continued)
Factors to consider
Select victim with least-negative effect on the system
Most common
Job priority under consideration: high-priority jobs
usually untouched
CPU time used by job: jobs close to completion usually
left alone
Number of other jobs affected if job selected as victim
Jobs modifying data: usually not selected for
termination (a database issue)

50

Starvation
Job execution prevented
Waiting for resources that never become available
Results from conservative resource allocation

Example
The dining philosophers by Dijkstra

Starvation avoidance
Implement algorithm tracking how long each job
waiting for resources (aging)
Block new jobs until starving jobs satisfied

51

Starvation (continued)

52

Starvation (continued)

53

Summary
Operating system
Dynamically allocates resources
Avoids deadlock and starvation

Four methods for dealing with deadlocks

Prevention, avoidance, detection, recovery

Prevention
Remove simultaneous occurrence of one or more
conditions
System will become deadlock-free
Prevention algorithms
Complex algorithms and high execution overhead
54

Summary (continued)
Avoid deadlocks
Clearly identify safe and unsafe states
Keep reserve resources to guarantee job completion
Disadvantage
System not fully utilized

No prevention support
System must detect and recover from deadlocks
Detection relies on selection of victim

55